Managing AI for end-users
Hi all, for enterprises on E5, from an administration and governance perspective:
- how do you allow end-users to use Copilot Studio, Power Platform, Power Automate with AI Builder, AI Models, and use of models from OpenAI or Anthropic?
- How do you manage the credits, allowing only specific users to certain features?
- Where do we find the “do-not-train-on-my-data-policy” if we select openai or anthropic models? Can we assume that data stays in our tenant? I don’t find it clearly written anywhere.
Edit: I’m very familiar with the OpenAI APIs. However Microsoft is inserting it all over its products, and introducing lots of abstractions with overlapping or ambiguous product and feature names, it is so hard to administer and also prescribe to internal users which features to use. In fact, it is somewhat easier to teach some users how to directly use the openai APIs — at least the token accounting and logging is clear.
Thanks for your help!